Despite the overwhelming evidence that shows the persistence of intra-annual variations on demographic events (deaths, birth dates and migration flows), life tables are computed and provided on an annual basis. This paper develops a new estimator for estimating sub-annual death rates that, considering the exact moment of occurrence (exact age and day) of events, concurrently accounts for ageing and calendar fluctuations. This paper also shows how modelling the intra-annual variations of death rates, through specific seasonal–ageing indexes, can be used as a tool for constructing new sub-annual tables from annual tables. This new methodology is exemplified using a real database of Spain made up of 186 million demographic events (1.5 million of which are deaths), from which seasonal–ageing indexes are estimated and conclusions drawn. First, seasonal effects are, as a rule, stronger than ageing effects. For a given integer age, season has a higher impact on increasing or decreasing the average risk of death at that age than the actual age of the exposed-to-risk. Second, the intensity of the effects varies among seasons and age-quarters. Third, neither seasonal nor ageing effects are age-stationary. Their impact, be it to varying degrees, intensifies as people get older. Fourth, there is interaction between seasonal and ageing effects. In short, life expectancies and probabilities of dying/surviving not only depend on people's age, but also on when their birthday falls within the year. This has implications, for instance, in managing pension systems or for insurance companies.